Theory of constraints

Niek,



I think i agree with you,

It’s harder and harder to get the right information to setup your schedule. so When 30% of the durations in your schedule is "estimated" by your self because otherwise your couldn’t complete your schedule on time. You should always ask your self are the critical paths primavera shows me, realy the critical paths. Yes i agree for all methods CPM, CCPM etc you should realy have the right basis information otherwise its no use then its just talk for sales and managers



Regards



Henk

We use other terminology. Spider Project calculates Resource Critical Path since 1992. In 1997 Goldratt called it Critical Chain. When simulating risks we manage contingency reserves that CC calls project buffer. Our contingency reserves are created for time, cost, material consumption - any parameters that are subjects to uncertainties and risks. This technology is widely used for projects of any size from tiny projects consisting of 500 activities to large models consisting of 92000 activities (this is not a restriction - just practical example). No problems with monitoring progress and anything else.

I think that Critical Chain and other approaches to scheduling shall be discussed in different forum.

Best Regards,

Vladimir

Well, that brings another interesting point into the discussion: how do you capture the actual performance on assigments?



If a project gets really large you’ll need electronic timehseets to capture the actuals. In Primavera this machanism work really well because it is based on a solid CPM based architecture. If you start introducing all kinds of buffers things can get really complicated.



Think about incorporating a change request in a CCPM schedule of a project in progress that uses assigment based time tracking.



Boy I’m I glad it’s weekend now...

Critical Chain Project Management (CCPM) relies on a fully resource loaded (CPM) schedule to establish the critical chain. Then it applies some common sense (like trying to avoid multi-tasking, account for Parkinson’s law, insert buffers where they matter, etc.) to enhance the quality of the schedule. Basically this is just another step on top of a CPM schedule that may make sense in certain situations (but certainly not always).



Now back to reality: only a handfull of people really know how to construct a proper CPM schedule, which accurately reflects the work to be executed. So throwing sophisticated probability theory and tools to it will only help if the foundation is right. Probably 90% of the CCPM implementation effort will be to establish solid CPM schedules before it makes any sense to apply CCPM techniques (same problem as with MonteCarlo simulation).



So with all respect for trends, I think people should learn to walk before they try to run, let alone try to fly.



Finally you are faced with a problem if you try to implement this in P3e because, any Earned Value Analysis will go banana’s on the project/resource/drum buffers used in CCPM. Furthermore this will have quite an effect on things like: ETC’s, resource forecasts, leveling, resource curves, etc.



Based on the above I don’t think we should expect a company like Primavera just adopting this trend but only add bits and pieces that would enhance the product over time.



Hope this helps,



Niek.

There is a previous Critical Chain Vs. CPM discussion. I’m sure there were a few others, but I can’t seem to locate them at the moment.



IMO, CC/TOC is mostly used in the IT/software space. I don’t see too much of it in aerospace, AEC or petrochem. YMMV.



Bernard Ertl

eTaskMaker Project Planning Software